A 3-D image of a cement paste with a resolution of 1 micrometer.
An image of the same paste that has been virtually hydrated for several hours.

In eight years, the project has progressed from version 1.0 to 7.1. The VCCTL is a challenging undertaking because its predictions are intentionally based on fundamental materials science and not empirical correlations. Then the VCCTL can be flexible and accurate enough to accommodate changes in materials that come about for various reasons, such as changes in supply or standards. Although some in the consortium have used the software already, NIST expects to have a more widely usable and practical version for producers in a few years.

There are two forces. Researchers are developing a technology that reduces routine testing that occurs in the mix design process. This is important in the material development process of lab mixes.

Second, when the mix design is established, VCCTL technology will make the remaining testing “smarter” by promoting the measurement of fundamental material quantities these kinds of models need to unleash their predictive power. An example for a portland cement concrete test would be measuring fundamental rheological quantities with a rheometer instead of running ad infinitum empirical slump tests.


Some ways the industry can use VCCTL include: selecting a cement and some mineral admixtures, along with certain kinds of fine and coarse aggregates, and predicting concrete rheology (workability), setting time, and temperature rise under specific curing conditions; predicting how green components of a cement binder will influence early-age properties, service life, and sustainability; and interpreting unusual test results and troubleshooting problems in the mix development process.

However, unlocking the power of fundamental models requires a higher level of material characterization than has routinely been performed before. The VC-CTL cement database is a good example of this kind of prerequisite information. In this database, we not only have the cement composition (mass fractions of each mineral component), but researchers also have the spatial distribution of those minerals in the cement powder, the readily soluble alkali fractions, the particle size distribution, and the actual 3-D particle shapes recorded mathematically.

Using this, they can reconstruct cement pastes with prescribed water-cement ratio, composition, volume and surface area fractions, and real particle shapes. The previous page shows a 3-D image of such a paste with a resolution of 1 micrometer, and the image to the right shows the same paste that has been “virtually hydrated” for several hours.

The combination of reality and model gives researchers an extraordinary ability to predict real behavior for portland cement concrete and someday for more complex formulations that include supplementary cementitious materials and admixtures.

NIST researchers are beginning to realize the potential of applying computational materials science to construction materials. Materials scientists worldwide are making great strides in understanding the fundamental physics and chemistry that govern the behavior of cement and concrete.

As these new insights are used to construct more accurate and sophisticated computer models, and as these modeling tools become more widely used, concrete materials research, testing, and quality control will change dramatically.

Edward Garboczi is Leader, Inorganic Materials Group; Jeff Bullard is a materials scientist in that group at the NIST. Visit Copyright 2009 Hanley Wood. No claims to original U.S. government works.